‘Over the Rainbow’ (Singing as Storytelling 1)

This is the first of a series of blog posts that I am calling Singing as Storytelling. When I was starting out as a singer, my first singing teacher told me: “If you’re lucky, the people who listen to you will say, ‘What a great voice!’, and ‘What a great musician!’. And if you’re really lucky, one day they will stop saying that, and they’ll start talking about the stories you tell, and the characters you bring to life, and how your performances change the listener.” Singers have to be good musicians, and they have to have solid vocal technique, but they become artists only when they truly understand lyrics and become (depending on the content of the lyric) either a storyteller or a character. These blogs introduce what I think are some outstanding examples of this.

‘Over the Rainbow’ started as a song in the 1939 film ‘The Wizard of Oz’ (although, amazingly, it almost didn’t make it into the final cut when the movie was released!). It won the Oscar that year for being the Best Original Song. The song works within the story of the film, but it also works independently of the musical, as a song about universal yearning for better times and a safer world – the reason it was adopted by the American armed forces in World War II. It was written by two Jewish men, sons of immigrants to the USA: Hyman Arluck (composer) and Isidore Hochberg (lyricist), better known as Harold Arlen and E.Y. (‘Yip’) Harburg. The song captures the mood of the Jewish people, on the eve of the Holocaust, yearning for a land they could call home, where the ‘clouds’ of war and persecution would be ‘far behind’, where ‘troubles melt’, away from the ‘chimney tops’ of the gas chambers and incinerators of the Nazis.

Here is the first outing of ‘Over The Rainbow’, sung by Judy Garland, aged 17, in the original 1939 movie. It is sweet, well sung and musical. But … it doesn’t plumb the depths and nuances of the lyric …

The film – and the song – catapaulted Judy Garland to stardom, and it became her signature song. Here she is singing it again, aged 33, in a TV special in 1955. Notice the extraordinary depth, subtelty and humanity she has developed in her interpretation 16 years after she first performed the song. (Don’t be put off by the constume and make-up – she performs this song straight after another number where she had to play a a tramp from the film ‘Easter Parade’!)

For me, this 1955 performance by Garland is the definitive version in terms of storytelling and character, and delivering the words.

Eva Cassidy did a memorable recording of this in 1992. This performance comes from January 1996; like Garland, she was 33 when she performed this (10 months before she died of cancer). Musically, it is a very sensitive and poignant performance. However, for me, it is makes a mess of the lyrics. For some reason, Cassidy deliberately breaks up the lines in odd places, and repositions weak syllables and unimportant words on climactic notes and downbeats. If this was an instrumental solo, it would be fabulous, but Cassidy has disregarded the genius of its composer and lyricist in her rendition of it as a song lyric, which I find frustrating. Even though I know the lyrics well, whenever I listen to Cassidy’s version, I get to the end of the song, and realise that I have not remotely been able to follow the emotional narrative of the song because of this disregard for lyrics, and their alignment with the contours of Arlen’s original melody.

Another very well known modern performance was by Israel Kamakawiwo’ole in 1988 (later recorded for his 1993 album with ukelele). The track has a hypnotic quality, but again, like with Cassidy’s version, there is nothing in the use of voice that suggests any appreciation for what the song is about, and Arlen’s beautifully designed melodic contours are put to one side in favour of Kamakawiwo’ole’s own, invented melody based on the chord sequence. Like with Cassidy’s version, I don’t follow the journey taken in the lyrics.


Here is another performance of the song, by Pink at the 2014 Oscars (the 75th anniversary of the song’s Oscar success). I include Pink’s version here as an example of how, in my view, not to sing it. She breathes, bizarrely, between ‘some’ and ‘where’, and then, even more unforgiveably, between ‘o-‘ and ‘-ver’. Why?! She makes a nonsense of the lyrics, and, indeed, of the English language. (Begins at 2:00)

Listen to Judy Garland’s 1955 performance again. After these strange performances by Cassidy, Kamakawiwo’ole and Pink, Garland’s version is a relief. Suddenly, the song makes sense again. It has depth, drama, humanity, and purpose. The song matters as an expression of a universal truth about human yearning.


7 thoughts on “‘Over the Rainbow’ (Singing as Storytelling 1)”

  1. What a fantastic post. I really enjoyed your analysis of the different performances from the storytelling point of view — something I wish more people considered when listening. I have seen so many comparisons of different performances, particularly in musical theatre where storytelling (you would think!) is key, and yet this aspect is hardly ever given any significance in these discussions. For example, when the 2004 “Phantom of the Opera” film came out, a lot of Broadway fans panned the performances of the leads because they saw them as insufficiently strong vocally — and yet, to me, the discussions of who hits the high notes most convincingly felt completely beside the point. Every phrase in the film felt real to me, acted as if it was not “a big number” but a moment of character development. I would be curious to know whether it’s just my insufficiently developed musical ear that allows me to enjoy these performances as much as I do! An example: https://youtu.be/yfPLh_6ckzI

    1. The way to listen for the ‘storytelling’ skill is to close your eyes. The storyteller-singer must act ‘with the voice’. The ‘Phantom’ example that you gave with the link has good cinematography and the actors making good connection with each other. But that is all ‘visuals’. In terms of precision and change of thought and feeling, I don’t hear it in the singers’ voices. The voices are sweet, but lack depth of characterisation, of deep connection to the meaning of the words. I have more blog articles on storytelling-singing coming up over the next few weeks. If you subscribe to my blog you’ll receive a notification when they come out.

  2. I’ll be interested to see your future articles. You are absolutely right that there is a sense of insufficiency here, but it is intentional. I should perhaps have found a different example from the film, because in its interpretation of the characters, this particular duet is supposed to be played “against” the lyrics, to some extent. It’s the sweet but non-passionate part of the love triangle, and so carries a lot of dramatic irony. But purely in terms of singing the lyrics, I really enjoy the emphasis being placed on what feel to me like the right keywords. Unlike the stage version I saw, Christine doesn’t figuratively put her foot down on, eg. “say you !LOVE! me”, and the result is more wistful and kinder for it. (Perhaps a better example, where the lyrics are not ironic, would be the last 15 mins of the film, e.g. https://youtu.be/pXkEAQKNj0k or the title song).

    As a completely different example, which unfortunately is not widely represented online, the fantastic jazz/swing/klezmer band Monsieur Camembert have what I think is the best album of Leonard Cohen covers in existence, precisely because of the way every song tells its story. I like the originals, but MC make the music speak the poems in a way that feels new. This is not strictly about vocals, just the entire interpretation, but your Judy Garland example made me think of jazz and then of MC.

    1. Yes, the second ‘Phantom’ piece you linked to has more character in the singing. But when he sings ‘free her’ at 2:13, and then repeats it at 2:17, for me, that is the perfect example of how NOT to do it. His second ‘free her’ is sung with exactly the same tone and energy as the first. A repetition of text/music must never, in my view, mean repetition of the same energy shape and intention by the singer. Repetition of material is, in psychological terms, a NEW place, and must have a fresh and different treatment. There must be a new psychological reason to repeat the material. He makes the same mistake on ‘I love her’ in 2:22 and 2:25; not only is the repetition the same, but it has the same energy as ‘free her’. Then again, ‘Christine, Christine’ at 2:30 has the same problem.

  3. That’s true. Interestingly, Raoul (Patrick Wilson) was the only one of the three leads who was an experienced musical theatre singer, and the only one who didn’t draw the ire of the Broadway fans (the viciousness of the comments about the other two astounded me). Another Phantom example since I’ve opened that door now: this is halfway through “Point of no Return”, as Christine turns the tables on the Phantom — https://youtu.be/Y7SzY7AoivU The little growl on “consume” at 0:48 is probably one of the reasons I love this.

  4. Wow Alexander, yes spot on.

    You really have no idea what the song is about from the other versions (though I am fond of the ukulele one!).

    The later Garland one is even enhanced by the costume/outfit I feel.

Leave a Comment

Your email address will not be published.